[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20200711050800.GX597537@hirez.programming.kicks-ass.net>
Date: Sat, 11 Jul 2020 07:08:00 +0200
From: Peter Zijlstra <peterz@...radead.org>
To: Steven Rostedt <rostedt@...dmis.org>
Cc: x86@...nel.org, linux-kernel@...r.kernel.org, mhiramat@...nel.org,
bristot@...hat.com, jbaron@...mai.com,
torvalds@...ux-foundation.org, tglx@...utronix.de,
mingo@...nel.org, namit@...are.com, hpa@...or.com, luto@...nel.org,
ard.biesheuvel@...aro.org, jpoimboe@...hat.com,
pbonzini@...hat.com, mathieu.desnoyers@...icios.com,
linux@...musvillemoes.dk
Subject: Re: [PATCH v6 14/17] static_call: Handle tail-calls
On Fri, Jul 10, 2020 at 08:23:19PM -0400, Steven Rostedt wrote:
> On Fri, 10 Jul 2020 15:38:45 +0200
> Peter Zijlstra <peterz@...radead.org> wrote:
> > @@ -1639,6 +1647,10 @@ static int decode_sections(struct objtoo
> > if (ret)
> > return ret;
> >
> > + ret = read_static_call_tramps(file);
> > + if (ret)
> > + return ret;
> > +
> > ret = add_jump_destinations(file);
> > if (ret)
> > return ret;
> > @@ -1671,10 +1683,6 @@ static int decode_sections(struct objtoo
> > if (ret)
> > return ret;
> >
> > - ret = read_static_call_tramps(file);
> > - if (ret)
> > - return ret;
>
> Hmm, what's the reason for moving this above? Should we have a comment
> here if there's importance that read_static_call_trampoline() is done
> earlier?
I suppose comments is something objtool lacks more of.
The reason is that add_jump_destination() is the thing that does
tail-call detection, and if it wants to add static-call sites, it needs
to know about the trampolines.
Powered by blists - more mailing lists